round results in aggregate table results (pyspark)

Question

Hello how would I round this content of table outputted by this code.

from pyspark.sql.functions import *
exprs = {x: "sum" for x in data2.columns[:4]}
data2.groupBy("Species").agg(exprs).show()

I've tried

round(data2.groupBy("Species").agg(exprs),2).show() #not ok

data2.groupBy("Species").agg(exprs).show().round(2) # not ok

Paul · Accepted Answer · 2019-10-29T21:56:06.957

4

round only works on one column. So you have to call it for each column, e.g.

agg_cols = data2.columns[:4]
exprs = [sum(col(x)).alias(x) for x in agg_cols]
aggregated_df = data2.groupBy("Species").agg(*exprs)
aggregated_df.select(col("Species"), *[round(c, 2) for c in agg_cols]).show()

edited Oct 29 '19 at 21:56

answered Oct 29 '19 at 21:41

Paul

1,114
8
11

I've got this error `TypeError: Column is not iterable` – abdoulsn Oct 29 '19 at 21:44
In which line? Can you check that agg_cols is a list of strings? – Paul Oct 29 '19 at 21:50
Yes agg_cols is list of string. It come from this line. `aggregated_df = data2.groupBy("Species").agg(exprs) ` – abdoulsn Oct 29 '19 at 21:52
1

I've edited the answer. Could you try again please? – Paul Oct 29 '19 at 21:56

round results in aggregate table results (pyspark)

1 Answers1